High-voltage isolating switch overheating state predicting method based on PSO-SVM

Abstract

The invention relates to a high-voltage isolating switch overheating state predicting method based on a PSO-SVM. Multiple factors influencing heating of an isolating switch like the load current, therated current, the environment temperature, the pollution grade, the latest overhauling year and the running state of the isolating switch are considered, an artificial intelligence algorithm of the particle swarm optimization support vector machine is adopted, and a high-voltage isolating switch overheating state predicting model is built. During application, original data or a testing sample during running of the high-voltage isolating switch is input to the built prediction model, the isolating switch state is predicted so that the heating state of the high-voltage isolating switch can be accurately predicted, a targeted overhauling strategy is timely taken aiming at the running state of the isolating switch, and the occurrence of power failure accidents is reduced.

Description

technical field [0001] The invention relates to the field of electric power engineering, and more particularly, to a method for predicting the overheating state of a high-voltage isolation switch based on PSO-SVM. Background technique [0002] High-voltage isolating switch is an important equipment used in conjunction with circuit breakers in substations, transmission and distribution lines, and its operation quality is closely related to the safety and stability of the power system. In the long-term operation of the isolation switch, due to the external environment, mechanical wear, arc ablation and poor contact, its overheating defects often occur. In the operation site, most of the high-voltage isolation switches do not have temperature monitoring equipment to monitor their operating temperature. [0003] Usually, the operation and maintenance personnel use infrared temperature measurement to detect the operating temperature of the isolation switch during regular inspect...

AI Technical Summary

Problems solved by technology

However, the generalization ability of the neural network algorithm is insufficient, and it is difficult to obtain accurate diagnosis results when the learning samples are incomplete in actual engineering.

[0005

Images